This study is a secondary data analysis of data from two randomized controlled trials (RCTs) conducted among Latina breast cancer survivors. The purpose of both studies was to test the effects of a culturally adapted 8-10 week stress management intervention on psychosocial distress and quality of life. The intervention, called Nuevo Amanecer (A New Dawn), was culturally tailored for urban and rural Latina breast cancer survivors, with extensive formative research and community input. The first RCT was conducted in five urban counties in Northern California: Alameda, Contra Costa, San Francisco, San Mateo, and Santa Clara [24,25]. The second RCT was conducted in three rural counties in California: Imperial, Tulare, and Santa Cruz/Monterey (Salinas and Watsonville areas of these counties) [26]. The self-reported survey measures used to assess English fluency, patient engagement, quality of breast cancer care and information, and physical and emotional well-being were identical across studies. Detailed methods for both studies are available elsewhere [24-26]. The current analyses used aggregated baseline data from both studies.
Sample Characteristics
The sample for this analysis consists of 304 women enrolled in the two Nuevo Amanecer studies. Although both studies focused on Latinas with non-metastatic breast cancer and recruitment and data collection methods were identical, inclusion criteria differed slightly across studies. The first study occurred in 2011-2014 and was the first efficacy trial of the adapted intervention, so eligibility was restricted to women living in the five urban counties who had been diagnosed within the past year. For the second trial conducted in 2016-2018, the intervention was adapted based on formative work with rural Latina breast cancer survivors and community partners to be suitable also for rural Latinas regardless of time since diagnosis. In both studies, community cancer centers with close ties to oncology clinics and hospitals were the primary sources of recruitment. Community-based recruiters who worked with community organizations serving Latinos with cancer contacted women in person or on the phone, verified their eligibility, and conducted the baseline survey in person prior to randomization. The current analyses used de-identified data, therefore, was not deemed human subjects research.
Measures
Predictors of interest included limited English proficiency (LEP), difficulty engaging with doctors, and rural/urban study site. English proficiency was assessed with the question “How well do you speak English?” with responses of: 1=not at all, 2=poorly, 3=fairly well, 4=well, 5=very well. Participants who responded not at all, poorly, or fairly well to the English proficiency item were identified as LEP. We developed a 3-item scale assessing difficulty engaging with doctors. Women rated the difficulty they experienced asking doctors questions about cancer and cancer treatment, telling doctors what they want, and asking for an interpreter using a 4-level response set: 1=not at all difficult, 2=slightly difficult, 3=somewhat difficult, and 4=very difficult. We developed these items based on our prior research developing measures of the quality of interpersonal processes of care suitable for use among diverse patients, including Latinos [27, 28]. The scale was scored as the mean of non-missing items, with a possible range of 1-4, and higher scores indicating greater difficulty engaging with doctors. The Cronbach’s alpha for the difficulty engaging with doctors scale was 0.72 in the sample. Women from the first RCT resided in major metropolitan areas and were classified as urban, while women in the second RCT were from areas whose economies relied primarily on agribusiness and were classified as rural.
Outcomes of interest were patient-reported quality of breast cancer care/information and physical and emotional well-being. We developed a 2-item measure to assess perceived quality of breast cancer care and information. Women were asked to rate separately the medical care and information they received for their breast cancer using a 5-point scale: 1=poor, 2=fair, 3=good, 4=very good, and 5=excellent. The 2-item scale was scored as the mean of non-missing items, with a possible range of 1-5, and higher scores indicating higher quality of care. The Cronbach’s alpha for the quality of breast cancer care and information scale was 0.94.
We used the Functional Assessment of Cancer Therapy-Breast (FACT-B) scales as breast cancer-specific quality of life measures assessing physical and emotional well-being, which were available in Spanish [29]. These scales are common in cancer research and well-validated and were scored per the developer’s instructions. The Emotional Well-Being Scale Score ranged from 0 to 20. Cronbach’s alpha for the Emotional Well-Being Scale was 0.87. The Physical Well-Being Scale Score ranged from 0 to 24. Cronbach’s alpha for the Physical Well-Being scale was 0.90. For both scales, a higher score indicates better well-being.
In both studies, breast cancer diagnostic and treatment information were collected using similar chart review abstraction methods. These measures, which served as covariates, included age at baseline (continuous), years since initial diagnosis (< 1 year, 1-5 years, or > 5 years), education level (less than high school, completed high school, more than high school), breast cancer stage at diagnosis (stage 0, stage I, stage II, or stage III), surgery type (breast-conserving surgery, mastectomy, or none), and treatment type (both chemotherapy and radiation, only chemotherapy, only radiation, or none).
Statistical Analysis
Descriptive statistics were used to characterize the sample and test for rural-urban differences in sample characteristics. Bivariate and multivariable linear regression models were used to assess the effects of limited English proficiency (yes versus no), difficulty engaging with doctors (continuous), and urbanicity (rural versus urban) on each of the three outcomes of quality of breast cancer care and information (continuous), emotional well-being (continuous), and physical well-being (continuous), controlling for demographic and breast cancer-related characteristics.